Skip to main content Skip to navigation

High spatiotemporal apple crop water use mapping using drone imagery

View Print Version

Written by Abhilash K. Chandel1, Lav R. Khot1, R. Troy Peters1, Claudio O. Stöckle1, Steve Mantle2

1Department of Biological Systems Engineering, Washington State University, Pullman, WA, Walla Walla, WA


Evapotranspiration (ET) is the actual crop water use and its mapping at the tree level could be critical for identifying problem areas in the orchard and site-specific irrigation scheduling and management (Chandel et al., 2021). Conventionally, single point ET is estimated using generalized crop coefficients and weather data (Allen et al., 1998). Other ET estimation tools or approaches include soil water budget, soil moisture, canopy stomatal conductance, and eddy covariance flux measurements. However, those provide limited in-orchard spatial variability (Chandel et al., 2021). Satellite-based imagery with energy balance models can provide geospatial ET maps (Allen et al., 2007). These maps are limited due to spatiotemporal resolution (e.g. ~99 ft/pixel and 16 days for Landsat 7/8 satellites). Cloud cover is an additional challenge that leads to no or poor ET estimation using satellite-based imagery. Drone-based imagery is an alternative to offer on-demand crop ET characterization at high spatial resolution (up to mm/pixel). Such resolution is also useful to assess spatial crop water use variabilities and to focus on trees or specific rows. Therefore, our research team has successfully utilized and validated drone-based multispectral and thermal infrared imagery with the energy balance model for mapping geospatial ET of a modern apple orchard at 2.8 in/pixel.

Drone imagery and energy balance modeling

A satellite imagery driven energy balance model, Mapping ET at High Resolution with Internalized Calibration (METRIC, Allen et al., 2007), was modified for drone-based multispectral and thermal imagery inputs (Chandel et al., 2020; 2021). A reference R package “water” (Olmedo et al., 2016) was modified for this model implementation. The model performs internal energy balance calibration to identify extremely stressed (hot, typically bare soil) and non-stressed material (cold, well irrigated vegetation) pixels. This calibration compensates for uncertainties in crop and weather parameters. The model inputs are: (1) surface reflectance (2) temperature and (3) digital elevation model orthomosaics, (4) drone flight metadata (flight date and time, sun azimuth and elevation angles), and (5) localized weather data (solar radiation, air temperature, relative humidity, wind speed, and precipitation). Using these inputs the model computes (1) net radiation, (2) soil heat flux, and (3) sensible heat flux and then the (4) latent heat flux as a residue which is converted to daily ET for drone-imaged block.

Orchard evaluation

Five drone imaging campaigns were conducted over a commercial apple orchard (cv. Buckeye gala) in the 2020 growing season (92, 76, 41, 21 days before harvest and 5 days after harvest). The drone (Quadcopter, AgBOT, Aerial Technology International LLC., Wilsonville, OR) used a 5-band multispectral imaging sensor (RedEdge 3, MicaSense, Inc., Seattle, WA) and a thermal-infrared imaging sensor (Duo Pro R, FLIR Systems, Wilsonville, OR). Images acquired from these sensors were geotagged using data from a GPS receiver module on the drone. Geotagging is useful to stitch and create seamless orthomosaics of the imaged orchard block. The drone had a skyward facing downwelling light sensor (MicaSense, Inc., Seattle, WA) to embed the solar irradiance metadata in the multispectral images during flights. The flights were configured using a Mission Planner application (version 1.3.49, Ardupilot, USA) for an altitude of 328 ft (100 m) above ground level (AGL) to capture images at 2.8 in/pixel (Multispectral) and 5.1 in/pixel (thermal infrared). Weather data was acquired from an all-in-one weather sensor (ATMOS-41, METER Group, Pullman, WA) installed by the WSU AgWeatherNet at the orchard center and 16.4 ft above ground level.

Collected imagery snapshots were stitched in a photogrammetry and mapping software platform (Pix4D mapper, Pix4D, Inc., Lausanne, Switzerland) to obtain the seamless orthomosaics discussed above. These stitched orthomosaics and weather data were inputs to the modified METRIC model (hereafter referred as DM) to derive daily ET as the geospatial data product. Those estimates were contrasted with two standard approaches: (1) ET from satellite-METRIC approach (hereafter referred as SM, Allen et al., 2007) and (2) canopy transpiration (T) from standard crop-coefficients and Penman-Monteith approach (Kcb-PM, Allen et al., 1998).

Fig. 1. Drone-based multispectral and thermal imaging of a high-density modern apple orchard
Results and Discussion

Drone imagery-based model (DM) shows higher potential to map in-orchard variability of crop water use compared to satellite-based (SM) approach (Fig. 2). The DM estimated mean daily ET of 0.25 in/day was similar to (Relative deviation = 12.3%) SM approach (0.28 in/day). Estimates from the two approaches also had a strong correlation (r = 0.82, Fig. 3a). The SM approach slightly overestimated the ET possibly due to the spatial resolution of satellite imagery that could have restricted the identification of the hot and cold anchor pixels near the target site or under similar conditions (Chandel et al., 2020; 2021). Furthermore, the drone imagery-based model effectively mapped the spatial water use variability, which was 31% for the imaged block. Such high-resolution variability maps may enable growers to implement block- or row-scale irrigation management.

For the season-long imaging campaigns, drone imagery-based daily canopy transpiration (T) estimates were in the ranges of 0.17–0.31 in/day, similar to the Kcb-PM approach (0.18–0.34 in/day). The estimates from the two approaches also had strong correlation (r = 0.95) and low relative deviation (<10%). The relatively higher estimates from the Kcb-PM approach were due to the use of non-stressed and generalized crop coefficients with minimal adjustments for local conditions. Whereas drone-based imagery provides needed specificity by capturing the actual crop information (stressed/non-stressed). Also, Kcb-PM provides a single point estimate (no variability) and is limited towards block- or row-scale irrigation management. Overall, high spatiotemporal resolution imagery could help to map crop water use and pertinent variability down to the tree-level. Intermediate data products (NDVI, canopy temperature, tree row volume, etc.) may also help growers to make decisions on canopy and other orchard management activities.


Fig. 2. Crop evapotranspiration maps derived from energy balance model implemented with (a) satellite imagery (99 ft/pixel) and (b) drone-based imagery (2.8 in/pixel).

Fig. 3. Comparison between drone imagery derived (a) evapotranspiration with the standard satellite approach and (b) transpiration with the standard crop-coefficient approach.

This study was funded by USDA NIFA projects (WNP0745 and WNP0893) and Washington Tree Fruit Research Commission. The authors would like to thank the cooperator grower from Columbia Reach, Chiawana orchard for providing the apple orchard for this study. The authors would also like to thank Dr. Ines Hanrahan from Washington Tree Fruit Research Commission, Ms. Bernadita Sallato, Mr. Jake Schrader, and Mr. Gajanan Kothawade from Washington State University for their assistance in the data collection.


Allen, R.G., Pereira, L., Raes, D., Smith, M., 1998. Crop evapotranspiration—Guidelines for computing crop water requirements. Food and Agriculture Organization of the United Nations Irrigation and Drainage Paper 56.

Allen, R.G., Tasumi, M., Trezza, R., 2007. Satellite-based energy balance for mapping evapotranspiration with internalized calibration (METRIC)-Model. J. Irrig. Drain. Eng., 133, 380–394.

Chandel, A.K., Molaei, B., Khot, L.R., Peters, R.T., Stöckle, C.O., 2020. Small UAS-based multispectral and thermal infrared imagery driven energy balance model for high-resolution evapotranspiration estimation of irrigated field crops. Drones, 4(3), 52–69.

Chandel, A.K., Khot, L.R., Molaei, B., Peters, T.R., Stöckle, C.O., Jacoby, P.W., 2021. High spatiotemporal water use mapping of a surface and direct-root-zone irrigated vineyard using UAS based thermal and multispectral remote sensing. Remote Sens. 13(5), p.954.

Olmedo, G.F., Ortega-Farías, S., de la Fuente-Sáiz, D., Fonseca-Luego, D., Fuentes-Peñailillo, F., 2016. Water: Tools and Functions to Estimate Actual Evapotranspiration Using Land Surface Energy Balance Models in R. R. J., 8(2), 352–369.

Washington State University